# morphik-org/morphik-core

Open-source multimodal retrieval engine (Morphik Core). By Morphik — AI back office for skilled nursing & senior living (morphik.ai).

Repository: https://github.com/morphik-org/morphik-core
Canonical: https://ross.abutalabs.com/products/morphik-core
Homepage: https://morphik.ai/docs
Language: Python
License: NOASSERTION
License Family: other
Topics: artificial-intelligence, cache-augmented-generation, colpali, database, multimodal, rag, rules-based-ingestion, litellm
Last push: 2026-07-23T05:48:19+00:00

## Health v2 (maintenance only)
Score: 64/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 94, release rhythm 35, longevity 47
- inputs: {"age_days": 660, "days_push": 41, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: no_releases, no_license
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 3708, forks 325 (observed 2026-08-28T04:08:15.126206+00:00)

## What it is
Morphik Core is an open-source, source-available multimodal retrieval engine for building RAG applications over unstructured data like PDFs, videos, and visually rich documents. It provides document ingestion, multimodal embeddings (ColPali), vector storage, retrieval APIs, MCP support, and multi-tenant user/folder scoping via Python/TypeScript SDKs and a REST API.

## Use cases
- build a RAG pipeline over PDFs and visually rich documents
- search tables and charts inside scanned documents with multimodal embeddings
- ingest videos and PDFs into a searchable knowledge base
- add retrieval-augmented generation to an AI application without stitching together OCR, embeddings, and a vector DB
- build a multi-tenant AI app with per-user and per-folder data isolation
- expose a document knowledge base to MCP clients like Claude
- query documents with an LLM and get grounded answers

## When to choose
- you need production RAG over complex, visually rich documents where plain text extraction fails
- you want an all-in-one ingestion, embedding, storage, and retrieval platform instead of assembling separate tools
- you need multi-user/folder scoping for multi-tenant applications
- you want built-in MCP support to plug your knowledge base into AI agents

## When to avoid
- you only need a lightweight vector store and already have your own ingestion and embedding pipeline
- you require a permissively licensed OSS dependency (license is source-available, not standard OSI)
- you need fully offline air-gapped operation but rely on the hosted Morphik cloud API
- your data is purely structured/tabular relational data better served by a SQL database

## Facets
- artifact type: service
- maturity: active
- function: rag, vector-database, search-engine, nlp, mcp, llm-inference, pdf, etl
- domain: artificial-intelligence, large-language-models, databases, pdf
- platform: python, self-hosted, cross-platform
- tags: multimodal-retrieval, colpali, cache-augmented-generation, unstructured-data, knowledge-base, document-ingestion, source-available, multitenancy, retrieval-augmented-generation, search, natural-language-processing, ai-agents, docker, web-server

## Member repositories
- morphik-org/morphik-core (main) score 64

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:08:15.126206+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-29T18:30:57.684090+00:00, confidence not recorded.
  - readme: https://github.com/morphik-org/morphik-core (fetched 2026-08-28T04:08:15.126206+00:00, sha 02579ee5b95e)
  - homepage: https://morphik.ai/docs (fetched 2026-08-29T09:24:43.957934+00:00, sha d904e40f747f)
  - site_page: https://dev.morphik.ai/docs/knowledge-base/how-do-i-set-up-rag (fetched 2026-08-29T09:24:43.975256+00:00, sha 26f51e1cad04)
  - site_page: https://dev.morphik.ai/docs/api-reference/getting-started (fetched 2026-08-29T09:24:43.970971+00:00, sha 1ee92a3745d0)
  - site_page: https://dev.morphik.ai/docs/python-sdk/morphik (fetched 2026-08-29T09:24:43.972707+00:00, sha fe69ed52e467)
  - site_page: https://dev.morphik.ai/docs/getting-started (fetched 2026-08-29T09:24:43.977592+00:00, sha bc5c0e8fc7a4)
  - site_page: https://dev.morphik.ai/docs/concepts/naive-rag (fetched 2026-08-29T09:24:43.979453+00:00, sha 8c41d0e5088b)
  - site_page: https://dev.morphik.ai/docs/core-functions/ingest-file (fetched 2026-08-29T09:24:43.981330+00:00, sha 92d09623292f)
  - site_page: https://morphik.ai (fetched 2026-08-29T09:24:43.968837+00:00, sha 5f067e8e480a)
- Data as of 2026-08-30T08:39:29.467469+00:00.